fuj-management/go/internal/io/cache/filecache.go

// Package cache implements a Drive-modifiedTime-gated JSON file cache,
// mirroring legacy/scripts/cache_utils.py.
package cache

import (
	"context"
	"encoding/json"
	"fmt"
	"fuj-management/go/internal/io/drive"
	"os"
	"path/filepath"
	"sync"
	"time"
)

// DriveClient is the subset of drive.Client used by FileCache.
type DriveClient interface {
	ModifiedTime(ctx context.Context, fileID string) (string, error)
}

type cacheFile struct {
	ModifiedTime string          `json:"modifiedTime"`
	Data         json.RawMessage `json:"data"`
	CachedAt     string          `json:"cachedAt"`
}

// FileCache wraps a Drive client to gate JSON file caching on sheet modifiedTime.
//
// Two TTL knobs mirror legacy/scripts/cache_utils.py:
//   - ttl: if the cache file on disk is younger than this, skip the Drive check entirely.
//   - apiCheckTTL: debounces in-memory Drive API calls per sheet ID.
//
// Atomic writes: data is marshaled to a .tmp file then os.Rename'd.
// Cache files are compatible with Python's format:
//
//	{"modifiedTime":"…","data":…,"cachedAt":"…"}
type FileCache struct {
	drive       DriveClient
	dir         string
	sheetMap    map[string]string // cache key → Drive file ID
	ttl         time.Duration
	apiCheckTTL time.Duration
	mu          sync.Mutex
	lastChecked map[string]time.Time
}

// New creates a FileCache.
// sheetMap maps cache keys to Google Sheets/Drive file IDs (mirrors CACHE_SHEET_MAP in config).
func New(d DriveClient, dir string, sheetMap map[string]string, ttl, apiCheckTTL time.Duration) *FileCache {
	return &FileCache{
		drive:       d,
		dir:         dir,
		sheetMap:    sheetMap,
		ttl:         ttl,
		apiCheckTTL: apiCheckTTL,
		lastChecked: make(map[string]time.Time),
	}
}

// Get returns the cached value for cacheKey, calling fetch if the cache is stale.
// T must be JSON-marshalable.
func Get[T any](ctx context.Context, fc *FileCache, cacheKey string, fetch func(context.Context) (T, error)) (T, error) {
	sheetID := fc.sheetMap[cacheKey]
	if sheetID == "" {
		sheetID = cacheKey
	}
	cacheFilePath := filepath.Join(fc.dir, cacheKey+"_cache.json")

	currentModTime, err := fc.currentModifiedTime(ctx, sheetID, cacheFilePath)
	if err != nil {
		return *new(T), fmt.Errorf("cache: modifiedTime for %s: %w", cacheKey, err)
	}

	// Try cache hit
	if data, ok := readCache[T](cacheFilePath, currentModTime); ok {
		return data, nil
	}

	// Cache miss — fetch fresh data
	fresh, err := fetch(ctx)
	if err != nil {
		return *new(T), err
	}
	if err := writeCache(cacheFilePath, currentModTime, fresh); err != nil {
		// Non-fatal: log but don't fail the request
		_, _ = fmt.Fprintf(os.Stderr, "cache: write %s: %v\n", cacheKey, err)
	}
	return fresh, nil
}

// Flush deletes all *_cache.json files in the cache dir and resets in-memory state.
func (fc *FileCache) Flush() (int, error) {
	fc.mu.Lock()
	fc.lastChecked = make(map[string]time.Time)
	fc.mu.Unlock()

	pattern := filepath.Join(fc.dir, "*_cache.json")
	matches, err := filepath.Glob(pattern)
	if err != nil {
		return 0, err
	}
	for _, f := range matches {
		_ = os.Remove(f)
	}
	return len(matches), nil
}

// currentModifiedTime returns a stable string representing the current version
// of the sheet, using the in-memory + file-mtime TTL guards before hitting Drive.
// On Drive failure, falls back to a 5-minute bucket string (matching Python).
func (fc *FileCache) currentModifiedTime(ctx context.Context, sheetID, cacheFilePath string) (string, error) {
	now := time.Now()

	fc.mu.Lock()
	lastCheck := fc.lastChecked[sheetID]
	fc.mu.Unlock()

	// Guard 1: in-memory debounce — skip Drive if checked recently
	if fc.apiCheckTTL > 0 && now.Sub(lastCheck) < fc.apiCheckTTL {
		if mt, ok := readModifiedTime(cacheFilePath); ok {
			return mt, nil
		}
	}

	// Guard 2: cache file is young enough — trust the stored modifiedTime
	if fc.ttl > 0 {
		if info, err := os.Stat(cacheFilePath); err == nil {
			if now.Sub(info.ModTime()) < fc.ttl {
				if mt, ok := readModifiedTime(cacheFilePath); ok {
					fc.mu.Lock()
					fc.lastChecked[sheetID] = now
					fc.mu.Unlock()
					return mt, nil
				}
			}
		}
	}

	// Hit Drive API
	mt, err := fc.drive.ModifiedTime(ctx, sheetID)
	if err != nil {
		// Fallback: 5-minute bucket string, matches Python _fallback_ttl()
		bucket := time.Now().Unix() / 300
		return fmt.Sprintf("ttl-5m-%d", bucket), nil
	}
	fc.mu.Lock()
	fc.lastChecked[sheetID] = now
	fc.mu.Unlock()
	return mt, nil
}

func readModifiedTime(path string) (string, bool) {
	cf, ok := readCacheFile(path)
	if !ok {
		return "", false
	}
	return cf.ModifiedTime, cf.ModifiedTime != ""
}

func readCacheFile(path string) (cacheFile, bool) {
	b, err := os.ReadFile(path)
	if err != nil {
		return cacheFile{}, false
	}
	var cf cacheFile
	if err := json.Unmarshal(b, &cf); err != nil {
		return cacheFile{}, false
	}
	return cf, true
}

func readCache[T any](path, currentModTime string) (T, bool) {
	cf, ok := readCacheFile(path)
	if !ok || cf.ModifiedTime != currentModTime {
		return *new(T), false
	}
	var v T
	if err := json.Unmarshal(cf.Data, &v); err != nil {
		return *new(T), false
	}
	return v, true
}

func writeCache(path, modTime string, data any) error {
	raw, err := json.Marshal(data)
	if err != nil {
		return err
	}
	cf := cacheFile{
		ModifiedTime: modTime,
		Data:         json.RawMessage(raw),
		CachedAt:     time.Now().Format(time.RFC3339),
	}
	b, err := json.Marshal(cf)
	if err != nil {
		return err
	}
	if err := os.MkdirAll(filepath.Dir(path), 0o755); err != nil {
		return err
	}
	tmp := path + ".tmp"
	if err := os.WriteFile(tmp, b, 0o600); err != nil {
		return err
	}
	return os.Rename(tmp, path)
}

// Ensure *drive.Client satisfies DriveClient at compile time.
var _ DriveClient = (*drive.Client)(nil)